[With the new version of its Software-as-a-Service solution Senseye Predictive Maintenance, Siemens aims to make human-machine interaction and predictive maintenance more efficient. Proven machine learning functions are enhanced with generative AI for this purpose.

Senseye Predictive Maintenance uses artificial intelligence and machine learning to automatically generate models for the behavior of machines and maintenance staff, directing the attention and expertise of users to where they are most needed. Based on this foundation, generative AI functionality is now being introduced. This will help users to utilize existing knowledge from all their machines and systems and select the right course of action to increase the efficiency of maintenance staff.

Currently, machine and maintenance data is analyzed by machine learning algorithms. The platform presents notifications to users in static, self-contained cases. The conversational user interface (UI) of Senseye Predictive Maintenance offers high flexibility and collaboration. With little configuration effort, it enables an interactive dialog between the user, the AI and the maintenance experts. This streamlines the decision-making process and makes it more efficient and effective.

From predictive maintenance to prescriptive maintenance

The generative AI scans and groups the recorded cases in the app regardless of language. This allows it to search specifically for similar cases from the past and their solutions in order to provide context for current problems. It is also possible to process data from different maintenance programs. To ensure the security of customer data, all information is processed in a private cloud environment.

In order for generative AI to transform the data into actionable insights, data quality is only of limited importance. With little configuration effort, maintenance logs and notes on previous cases can also be taken into account. By better contextualizing the available information, the app is not only able to detect anomalies in the production process, but also to proactively derive a suitable maintenance strategy, so-called "prescriptive maintenance".